18: Sleep and Epilepsy


CHAPTER 18
Sleep and Epilepsy


Sofia Khan and Sofia H. Eriksson


STRUCTURE OF NORMAL SLEEP


Sleep is divided into stages, defined by changes in EEG, electro‐oculogram (EOG), and muscle tone (EMG). Using these measures, wakefulness, light sleep, slow‐wave sleep (SWS), and rapid eye movement (REM) sleep can be differentiated.


Sleep onset is characterised by a brief transitional phase (stage 1) in which the alert and wakeful alpha rhythm (8–13 Hz) starts slowing, and slow theta (4–7 Hz) waves appear. EOG usually demonstrates rolling eye movements. EMG activity may be reduced, but is still present. Stage 2 sleep is characterised by episodic bursts of central, 14–16 Hz rhythms (sleep spindles), and high‐amplitude predominately frontal biphasic waveforms (K complexes) in the EEG. EMG activity is similar to that in stage 1. In stage 3 sleep (slow wave sleep, SWS), the EEG is characterised by high‐amplitude (>75 μV) delta activity (0.5–4 Hz). SWS is associated with a slowing of cardiac and respiratory rates, lowered blood pressure, muscular relaxation, and a reduction in metabolic rate. During REM sleep, the EEG shows a desynchronised mix of theta and alpha waves, low voltage, and occasional bursts of sawtooth waves associated with rapid eye movements on EOG and atonia on EMG. Sleep parameters in an adult are summarised in Table 18.1.


Epilepsy and sleep are intertwined. Sleep affects epilepsy, with non‐REM sleep facilitating and REM sleep inhibiting epileptic activity.1 Sleep deprivation and sleep disorders may trigger seizures, and some seizure types typically occur from sleep or following awakening. Epilepsy and its treatment may exacerbate sleep disorders. Sleep complaints are twice as common in people with epilepsy than in controls. There may be differential diagnostic challenges for individuals experiencing nocturnal or sleep‐related events.


EPILEPSY AND CIRCADIAN RHYTHMS


Circadian patterns in seizures have been confirmed using chronic intracranial EEG recordings.2 Seizure occurrence is influenced by several factors, including epilepsy type (generalised or focal) and site of onset (e.g. frontal, temporal). In generalised epilepsies, seizures tend to occur following sleep. In focal epilepsies, frontal lobe seizures occur predominantly during sleep. Temporal lobe seizures typically occur during wakefulness, with a bimodal distribution: a primary peak in the late afternoon and a secondary peak in the morning.3 It is unclear if this is due to behavioural states (wakefulness vs sleep or drowsiness), environmental conditions, or independent effects of the endogenous circadian system. Support for an endogenous circadian distribution of seizures, irrespective of the sleep–wake status, has been shown in animal models. Variability in cortical excitability across the circadian cycle and following sleep deprivation has also been demonstrated by transcranial magnetic stimulation. Cortical excitability increases with time awake and varies according to epileptic syndrome, but is also modulated by the circadian phase, with lower cortical excitability in the evening. Variability in circadian seizure preponderance also opens the possibility for chronotherapy.4 This involves treating at times of most significant risk of seizures based on historically high seizure or epileptogenicity levels concerning wakefulness, sleep, circadian, or non‐circadian rhythms.


Table 18.1 Approximate normal values for sleep parameters in an adult.






















Sleep efficiency >90%
Stages 1&2 (light sleep) 50–70%
Stage 3 (slow wave sleep, also known as deep sleep) 15–20%
REM sleep 15–20%
Latency to sleep onset >10 minutes
Latency to REM sleep onset 60–90 minutes

DIFFERENTIAL DIAGNOSIS


Diagnosing the nature of sleep‐related events can be challenging. Several scales are available to differentiate epilepsy and parasomnias, particularly non‐REM parasomnias, for example, the frontal lobe epilepsy and parasomnia scale.5 The sensitivity and specificity of these scales are not well established and are rarely used in the clinic.


Four characteristics may facilitate differential diagnosis (Table 18.2).



  1. Timing of events during sleep – non‐REM parasomnias tend to occur in the first third of sleep time, when slow‐wave sleep is most prevalent. REM sleep behaviour disorder (RBD) typically occurs during the second half of sleep, when REM sleep is most pervasive.
  2. The number of events per night – non‐REM parasomnias tend to occur 1 – 3 times/night, while seizures may occur much more frequently.
  3. Frequency of events over time – non‐REM parasomnia events usually vary with time (depending on precipitating or exacerbating factors), while seizure and RBD occurrence may be more consistent.
  4. Age at onset and development of the disorder over time – non‐REM parasomnias often start in childhood, and many grow out of it in their teens. Still, adult‐onset does not exclude the diagnosis. Epilepsy can begin at any time of life and tend not to lessen unless treated with ASM. RBD tends to start later in life.

Epileptic seizures from sleep may have unusual features, causing diagnostic confusion. They can be misdiagnosed as parasomnias or dissociative seizures and may be drug‐resistant. Episodes are stereotyped, often frequent, brief, and may only occur during sleep. Events may include dystonic posturing and apnoea. They can be uni‐ or bilateral and may retain some awareness. They often arise from light sleep, whereas non‐REM parasomnias usually arise from N3 sleep. Frequency is highly variable; some people have more than 20 seizures per night. The occurrence of FBTCS and seizures in wakefulness helps secure the diagnosis.


Table 18.2 Features to help differentiate between NREM parasomnia, REM sleep behaviour disorder, and frontal lobe seizures.












































Feature NREM parasomnia REM sleep behaviour disorder Seizures
Age of onset Usually starts in childhood and tends to settle. Can re‐emerge in adult life Later in life, often >50–60 years.
Narcolepsy can occur in teenage years or early adult life		 Variable, often 1st or 2nd decade
Timing during sleep Often in the first 3rd of sleep Second half of the night Throughout the night
Frequency 1–3 times/night.
Stress/anxiety, sleep deprivation are typical triggers. Noise, touch, and alcohol are less common triggers.
Frequency varies, can have prolonged remission periods and periods with frequent events		 1–2 events/night. Usually, most or every night Variable, often in clusters, usually with a higher frequency of events than parasomnias (1–30/night)
Sleep stage events occur Deep NREM sleep (N3) REM sleep NREM sleep, often in light sleep and transitions between sleep stages
Duration 1–30 minutes A few seconds to 1 minute 10–60 seconds
Onset and offset Sudden onset but usually gradual offset, after which the person may remain confused or return to sleep Sudden onset and offset, and often return to sleep, movements can wake the person Sudden onset and offset. Often rapid recovery with little confusion
Semiology Eyes open. Variable complexity, not stereotyped. Can get up and leave the bed. Episodes vary and include night terrors, sleepwalking, and sexsomnia Eyes closed. Variable, often jerky movements and/or vocalisation. Do not get up and out of bed (may fall out of bed). May recall dreams congruent with movements Highly stereotyped, often with dystonic posturing or hypermotor features

The term nocturnal frontal lobe epilepsy was commonly used in the past. As the seizures are sleep‐related and may arise from extra‐frontal brain regions, the condition is now designated as sleep‐related hypermotor epilepsy.6 The individual may be woken by an aura, although this is often not recalled when events arise from sleep. Complex automatisms or confused postictal behaviour, during which the person gets out of bed and wanders around, can also be confused with parasomnias.


Other parasomnias/sleep‐related movements which can be misdiagnosed as seizures include:



  • RBD is less likely to cause diagnostic confusion than NREM parasomnias. RBD is characterised by loss of normal REM atonia, resulting in dream enactment behaviours.7 Episodes occur in the last half of the night when REM sleep is most prevalent, whereas seizures are more likely soon after sleep onset. In RBD, dreams are often, but not always, recalled, and the behaviour is congruent with the dream. Unlike seizures, they are not stereotyped, and the eyes are closed. A definitive diagnosis is made through polysomnography (PSG), which shows loss of atonia during REM sleep and associated dream enactment behaviours.
  • Periodic limb movements of sleep (PLMS) may be an age‐related phenomenon (less than 1% of young adults but occurring with increasing frequency during middle and old age). Movements occur at regular intervals of 5‐90 seconds, while seizure occurrence is irregular. PLMS occur mostly during light sleep but occasionally in quiet wakefulness. PLMS usually manifest as an extension of the big toe and ankle dorsiflexion, but can also cause hip, upper limb, or whole‐body movements, raising the differential of seizures. PSG is needed to diagnose PLMS.
  • Hypnic jerks commonly occur in healthy individuals upon falling asleep. The whole body can jerk. Episodes are more common when sleep‐deprived and anxious. Myoclonic jerks associated with generalised epilepsies tend to occur in the morning following awakening.
  • Rhythmic movement disorder (RMD) is characterised by repetitive, stereotyped rhythmic movements that involve large muscle groups (e.g. body rocking and head banging). Episodes are common in children and usually resolve by age of five, but can persist into adulthood. These occur during the transition from wakefulness to sleep but may continue in light sleep. Unlike seizures, episodes of RMD stop when the individual wakes.
Tags:
Oct 17, 2026 | Posted by in NEUROLOGY | Comments Off on 18: Sleep and Epilepsy

Full access? Get Clinical Tree
Get Clinical Tree app for offline access